The Unbearable Ambiguity of Emoji
A picture is said to be worth a thousand words, implying you’d need a thousand words to describe a single image. Emoji are much simpler—maybe just ten words would suffice. But there’s still plenty of room for interpretation within the picture/text exchange rate, as a recent Canadian court case shows.
A grain buyer in Saskatchewan, Canada, texted farmers asking to buy flax at $17 per bushel, and one farmer indicated he could deliver on that. The buyer texted a photo of the contract, as it had done numerous times before, and asked, “Please confirm flax contract.” The farmer replied with a thumbs-up 👍 emoji. But when prices for flax jumped to $41 per bushel and the farmer failed to deliver on the less-lucrative contract, the buyer sued for breach.
The facts of the case aren’t in question; the nut of the issue is whether the farmer’s 👍 response had the legal meaning of “I accept the contract that you just texted me” or merely implied “I’m acknowledging receipt of the contract.” Judge Timothy Keene ruled that the farmer’s 👍 counted as a signature and the farmer had thus breached the contract:
This court readily acknowledges that a 👍 emoji is a non-traditional means to “sign” a document but nevertheless under these circumstances this was a valid way to convey the two purposes of a “signature” – to identify the signator (Chris using his unique cell phone number) and as I have found above – to convey Achter’s acceptance of the flax contract.
This emoji-driven ruling didn’t come out of thin air. The grain buyer’s attorney said in documents filed in the case that the farmer had previously agreed to contract at least four other times in text by replying “looks good”, “ok”, or “yup”.
More generally, the farmer’s lawyer argued that:
…allowing a simple 👍 emoji to signify identity and acceptance would open up the flood gates to allow all sorts of cases coming forward asking for interpretations as to what various different emojis mean – for example what does a 👊 emoji mean or a 🤝 emoji mean, etc.
The judge wasn’t swayed, responding:
This appears to be a sort of public policy argument. I agree that this case is novel (at least in Saskatchewan) but nevertheless this Court cannot (nor should it) attempt to stem the tide of technology and common usage – this appears to be the new reality in Canadian society and courts will have to be ready to meet the new challenges that may arise from the use of emojis and the like.
This story has been making the rounds based primarily on its humor value, often missing the fact that the farmer had laconically agreed to contracts before. But I’ve long been perturbed about the use of emoji in text communications when there’s room for interpretation. It’s one thing to toss in a smiley or frowning face to add emotional color to text, but entirely another to rely on an image to convey something significant. Sometimes a cigar is just a cigar, but an 🍆 isn’t always just an eggplant.
It was in this context that I read about the draft Emoji 15.1 list that contains recommendations for new emoji that will likely become final in September 2023, alongside the release of Unicode 15.1. (Apple would likely integrate them into its operating systems in early 2024.) This batch of emoji is very particular—most are new sequences that combine existing characters. Of the 118 candidates, 108 merely change directionality, so a slew of emoji that currently only face left will now be available facing right as well. That seems uncontroversial, but meaning still rears its ugly head with the proposed Head Shaking Horizontally and Head Shaking Vertically emoji. Emojipedia says:
Designs for the 🙂↔️ Head Shaking Horizontally and 🙂↕️ Head Shaking Vertically may pose some semantic difficulties for emoji designers, given that they can have inverted meanings in different cultures. For example, while a nodding head such as intended to be depicted in the 🙂↕️ Head Shaking Vertically emoji is a positive “yes” in the United States, in Bulgaria it conveys a negative “no” meaning.
There’s no stuffing emoji back into the bottle, though I suspect relatively few TidBITS readers use emoji heavily to stand in for true textual communications, rather than as simple reactions or emotional color. Even if you know what you intend an emoji to mean, there’s no telling how someone else will interpret it. And the next time you’re selling flax, perhaps stick with saying what you mean in actual words until the powers that be develop a Signature emoji.
I was pretty bothered when I read this news as well. I don’t have a solution, but a few points:
Normally, I would say one should never conduct business by text message. Have your negotiations via voice call and require an actual signed document before the deal is considered concluded.
FWIW, the landscape company I deal with works almost along these lines. We will discuss things on the phone, but before any work is done, the company sends me a link to an e-signing web site with the full text of the contract. I have to e-sign it (including providing a credit card to pay the deposit) before any work begins. Then on completion, I go back to the site to pay the balance.
But in this case, it may be a bit different, because we’re talking about commodity exchange activities where (I assume) it is customary to make legally binding decisions on the basis of quick phone calls. So a text might not be so far out of line.
The fact that this particular farmer has previously approved contracts is very significant here. He’s got a reputation for approving contracts with ambiguous language, and has previously executed them based on them.
So it begs the question: Would he have voiced the same opinion if the price of flax hadn’t risen so dramatically between his approval and execution? I suspect that he would not have, and was trying to use his ambiguous response as a way to get out of a deal that in hindsight proved to be less lucrative than it might otherwise have been.
But for the rest of us, the lesson is clear: Leave the emojis and casual language to communication with friends and family. Keep it nice and formal where business is concerned.
Contract law has a principle that ambiguity works in favor of the party that did not draft the ambiguity. (If anyone is interested, there was a decision that cost Rogers Communication an estimated two million dollars, presumably Canadian, because of a comma.) To that extent, it seems that the court ruling went the right direction.
Yes, and proofread.
An interesting article, but I’m not sure I agree with the premise that the use of an emoji was in any way problematic. It’s pretty clear what the
meant (as the court found), and I can’t see that it’s any more ambiguous than ‘yup’ or ‘ok’. In some ways, I see it as less ambiguous.
Yes, emojis can be ambiguous and mean different things to different cultures. But so can words, particularly colloquial ones. I’d be interested if there is any research into whether emoji are any worse or better in this respect, because my sense is a lot of the discomfort with emoji is based on anecdotal evidence and an older generation (of which I am a part of!) that didn’t grow up with them.
Anecdotally, I’ve found emoji decrease the ambiguousness of message-based communication. It is often hard to convey tone in short bits of text, and a carefully chosen emoji (like carefully chosen punctuation) can make all the difference. Yes, the wrong or misunderstood emoji can cause embarrassment or offence, but that is no different from text (and verbal) communication more generally.
I see handwringing about how inappropriate or even dangerous emoji use is for certain forms of communication because it’s not ‘proper’ and can be misinterpreted. But I haven’t seen any proper evidence and my experience certainly doesn’t bear it out. And if anything the article linked in this piece shows the opposite. It feels very much to me like lawyers who oppose the use of any punctation in sentences because it could introduce ambiguity, when all it does is make things hard to read!
There’s definitely research into this, for instance showing that different platform renderings of the same emoji can result in different interpretations.
and some of the same people looked into how much less misinterpretation happens when emoji are embedded in contextual text.
Interestingly, that paper suggests that old-style emoticons have fewer issues with interpretation, perhaps because there were so few regularly used.
That’s interesting, thanks for the link. However, I think it is too outdated to be applicable in 2023 – there’s been a huge amount of standardisation in emoji representation across platforms since that research was carried out. For instance, if you look at the emoji they call out in the article, it is now consistent across the board (with the possible exception of the Softbank version, though the blushing cheeks probably make it close enough):
As you can tell, I’m still sceptical that this is an issue in practical terms
OK, here’s a more recent summary of a paper talking about how the meaning of the same emoji can change over time.
Actual paper, with lots of references to other research on emoji:
And an interactive tool so you can see how meaning has changed for any given emoji from 2012–2018.
https://emoji-semantic-change.herokuapp.com/
That’s very interesting, thanks for pointing me to it – the interactive tool is great! I like how they present the four characteristic patterns of change. It makes sense, and helps me understand the model of how different types of emoji change meaning.
In terms of the issue that this thread is about, the paper seems to be putting forward the notion that emoji are very much like words in how they are used and understood.
Though some emoji clearly have had shifts in meaning (the four characteristic patterns they show). Doesn’t this support the idea that emoji aren’t exceptional in their potential for misunderstanding? I think what I’ve yet to understand is what makes emoji different from our language and communication more generally, especially as new words and phrases enter it?
I did have a look through the references in the paper, but most didn’t consider whether emoji are likely to cause misunderstandings. Those that did are from circa 2015/2016, before emoji were standardised across platforms. None were from a linguistics perspective considering if emoji have characteristics that differ from spelled words in terms of their proclivity for ambiguity or likelihood to be misunderstood.
One thought I just had about that research is that it’s looking at the changes in meaning of emoji in usage on Twitter from 2012 to 2018. What that misses is that the people who either don’t understand the meaning of an emoji or disagree with that meaning aren’t going to use it at all. I almost never use emoji, and when I do, only a very small number of them in uncontroversial ways, because I can’t say that I even know how they might be interpreted.
If you think back to the discussion we had about which emoji to provide here in Discourse, it became very clear that people really didn’t agree about quite a few of the possibilities.
I’ve been intrigued with the ambiguity of emoji (is it emojis??) and that recent headline.
I correspond with friends who are from different cultures from mine. Even for those acculturated here, at least for some humor, I’ll disambiguate, as for example:
”In the American sense, not the European.”
“As in ‘hearty agreement,’ rather than “Hook ‘em Horns!,’ or “cuckold.”
Or, as a proud graduate of University of Texas Medical Branch,
It would be fun to hear other people’s take on ambiguous emoj, the most classic being “chocolate ice cream”
being interpreteted as “fecal material.”
This article should be titled: The Unbearable Stupidity of People. What else can be said when people—including emoji users, emoji recipients and judges who decide emoji law suits—don’t have a brain in their bodies?
Yes, good point. And additionally, whilst it provides a wealth of real-world usage information, Twitter is also not necessarily a representative sample of the population generally. So there is a lot of usage and non-usage left out. Hopefully over coming years more linguistic-based research is carried out. I find word usage and how we communicate (and how it changes) fascinating.